Journal of Biomechanics 45 (2012) 427–433

Contents lists available at SciVerse ScienceDirect

Journal of Biomechanics
journal homepage: www.elsevier.com/locate/jbiomech
www.JBiomech.com

Review

Muscle fatigue – from motor units to clinical symptoms

Roger M. Enoka n
Department of Integrative Physiology, University of Colorado, Boulder, CO 80309-0354, USA

a r t i c l e i n f o abstract

Article history: Reductionist approaches have provided little insight on the fatigue experienced by humans during
Accepted 16 November 2011 activities of daily living. Some of the reasons for this lack of progress include the persistence of
outdated concepts, the misinterpretation of experimental recordings, and a failure to embrace a global
Keywords: perspective on fatigue. This paper summarizes the three examples of these limitations that were
Muybridge discussed in the 2011 Muybridge Award lecture: motor unit types and muscle fatigue, myoelectric
Marey manifestations of fatigue, and fatigue and fatigability. Although the motor units in a population do
Fatigue exhibit a range of fatigability values, there are not distinct groups of motor units and the concept that
Fatigability some motor units are resistant to fatigue emerged from protocols in which motor units were activated
Perceptions of fatigue
by electrical stimulation rather than voluntary activation. The concept of distinct motor unit types
Motor unit
should be abandoned. The second example discussed in the lecture was the use of surface EMG signals
Myoelectric
Electromyogram to assess fatigue-related adjustments in motor unit activity. The critical assumption with this approach
Muscle force is that the association between surface EMG amplitude and muscle force remains constant during
Computational model fatiguing contractions. Unfortunately, the relation does not remain constant and a series of computa-
tional studies demonstrate the magnitude of the discrepancy, including the absence of an association
with the activation signal emerging from the spinal cord and that received by the muscle. The third
example concerned the concepts of fatigue and fatigability. It has long been recognized that fatigue
involves both sensations and impairments in motor function, and the final part of the lecture urged the
integration of the two constructs into a single scheme in which fatigue can be modulated either
independently or by interactions between perceptions of fatigue and the mechanisms that establish
levels of fatigability. The expectation is that such critical evaluations of the concepts and approaches to
the study of fatigue will provide a more effective foundation from which to identify the factors that
contribute to fatigue in health and disease.
& 2011 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
2. Motor unit types and fatigue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
3. Myoelectric manifestations of fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
3.1. EMG–force relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
3.2. Fatigue induced depression of EMG amplitude. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
4. Fatigue and fatigability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Conflict of interest statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

1. Introduction

On the occasion of the 2011 Congress for the International

Society of Biomechanics held in Brussels, I was recognized for career
achievements in biomechanics and honored with the prestigious
n
Tel.: þ1 303 492 7232; fax: þ1 303 492 6778. Muybridge Award. In recognition of the honor, I dutifully began the
E-mail address: enoka@colorado.edu Muybridge Lecture by acknowledging the contributions to the field

0021-9290/$ - see front matter & 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jbiomech.2011.11.047
428 R.M. Enoka / Journal of Biomechanics 45 (2012) 427–433
of biomechanics of the pioneer after whom the award was named,
Eadward James Muybridge (1830–1904). The Society describes
Muybridge as the ‘‘father of cinematography’’. Although he did
accumulate 20,000 images of animal and human locomotion over
an 8yr period, he was more interested in the representation of
motion from an artistic perspective than in analyzing its attributes.
Rather, it was a contemporary of his who had the same initials and
was born and died in the same years who contributed more than
Muybridge to the scientific study of human motion. His name was
Etienne-Jules Marey (1830–1904). The substantial contributions to
the study of human motion by the venerable professor at the Collége
de France and member of the Academy of Sciences are described in
an excellent biography written by Marta Braun (1992). Marey and
Muybridge were aware of each other’s work, and their relatively
brief interaction is described in the proceedings of a conference to
celebrate the 100 yr anniversary of the beginning of cinema
(Delimata, 1996). Accordingly, the lecture was dedicated to EJM
(1830–1904).
The purpose of the lecture was not to provide an historical
account of key developments in biomechanics, but rather to
discuss the difficulties associated with changing the ideas that
define a contemporary issue. The topic chosen for the occasion
was muscle fatigue, and the approach was to proceed from ideas
on muscle fatigue that should be discarded through to challenges
for the future. This was accomplished by discussing motor unit
types and fatigue, myoelectric manifestations of fatigue, and the
distinction between fatigue and fatigability.
2. Motor unit types and fatigue
The motor unit denotes the basic functional element of the
central nervous system and muscle that produces movement. It
comprises a motor neuron in the ventral horn of the spinal cord,
its axon, and the muscle fibers that the axon innervates
(Duchteau and Enoka, 2011; Sherrington, 1925). The central
nervous system controls muscle force by varying the activity of
the motor units in the muscle. The force exerted by each motor
unit depends principally on the number of muscle fibers that are
innervated by the motor neuron and the rate at which the motor
neuron discharges action potentials. The motor unit population
that innervates a muscle is heterogeneous, due to systematic
variations in the properties of both the motor neurons and the
muscle fibers. These systematic variations in the components of
the motor unit have been used to distinguish different types of
motor units.
Burke and colleagues developed the classic scheme for distin-
guishing motor unit types. In a study on experimental animals,
single motor neurons were activated with brief electric stimuli
and the force exerted in the muscle fibers innervated by the
motor neuron was measured (Burke et al., 1973). They found that
two contractile properties could be used to identify three types of
motor units; the contractile properties were the profile of a
submaximal tetanus and a measure of fatigability (Fig. 1). The
protocol began by recording the twitch response of the motor unit
to a single electrical stimulus and measuring the time from the
onset of the increase in force to the peak force, which is a measure
referred to as contraction time. Next, the motor unit was acti-
vated with a series of electric stimuli with an interstimulus
interval of 1.25  the contraction time of the unit. The resulting
tetanus either exhibited a progressive increase in force or there
was a slight decline that began soon after an initial peak force
(Fig. 1). The decline in the tetanic force is known as the sag
property. Burke and colleagues classified motor units that exhib-
ited sag as fast (type F) units, whereas those that did not display
sag were referred to as slow (type S) units. Sag appears to be
Twitch
1.25x CT
Type F Type S
40 Hz for 330 ms at 1/s
0s 0s
120 s
30 s
300 s
60 s
Type FF Type FR
Fig. 1. The protocol used by Burke et al. (1973) to classify motor units in the cat
gastrocnemius muscle into three categories based on contractile properties. The
protocol comprised three steps: (1) recording the twitch response to a single
electrical stimulus applied to the motor axon and measuring the time to reach
peak force (contraction time, CT); (2) delivering a train of stimuli with an interval
of 1.25  CT between successive stimuli and observing whether or not the tetanus
exhibited sag; and (3) administering a fatigue test that involved measuring the
decline in force in response to repetitive trains of 40 Hz stimulation. Those motor
units that exhibited sag were classified as type F units and then were character-
ized as being fatigue sensitive (type FF) or fatigue resistant (type FR) based on the
decline in tetanic force during the fatigue test.
caused by a transient reduction in the duration of the contractile
state (Carp et al., 1999).
The other classification property was fatigability, which was
assessed as the decline in the peak force in response to repeated
activation of the muscle fibers with trains of electric stimuli. The
fatigue test involved repeated submaximal tetani that were
evoked once each second for several minutes. Each tetanus
comprised the force elicited by 13 stimuli at 40 Hz for 330 ms.
The tetanic force elicited by each 330 ms train of stimuli either
declined minimally or substantially over the course of the fatigue
test (Fig. 1). Those units whose tetanic force did not decline or
decreased only slightly were described as fatigue resistant. In
contrast, motor units that exhibited a marked decline in the
tetanic force were characterized as fatigue sensitive or fatigable.
The decline in tetanic force was quantified with a fatigue index,
which was calculated as the force after 120 s relative to the initial
force. The fatigue index for the fatigue resistant units was Z0.75,
whereas it was o0.25 for the fatigable units.
On the basis of the sag and fatigue tests, Burke et al. (1973)
identified three types of motor units: (1) type S–very fatigue
resistant units with relatively slow twitch contraction; (2) type FR
– fatigue resistant units with fast twitch contraction; and (3) type
FF – fatigue sensitive units with relatively fast twitch contraction.
Unfortunately, these descriptions are somewhat misleading
because the type S and F units were distinguished on the basis
of sag and not twitch contraction time. Indeed, the continuous
distribution of twitch contraction times (Fig. 2) indicates that
there were not distinct groups of slow- and fast-twitch motor
units in the cat gastrocnemius muscle. Similarly, the time to peak
force for motor units in the human tibialis anterior muscle as
estimated with spike-triggered averaging exhibit a relatively
normal distribution and argue strongly against the existence of
distinct groups of slow- and fast-twitch motor units (Fig. 2).
The other significant constraint of the motor unit-typing
scheme is the measure of fatigability. As indicated by Burke
 R.M. Enoka / Journal of Biomechanics 45 (2012) 427–433
15
12
9 were discharged. Furthermore, the recruitment thresholds of the
Count
6 units contributed less to the net muscle force as the task
3 tary contractions.
0 and that the characterization of motor unit fatigability with tests
20 50 80
Contraction time (ms)
0.2
Proportion of motor units
0.15
0.1
0.05
0 reflected in the magnitude of the activation signal discharged by
20 30 40 50 60 70 80
Time to peak (ms)
Fig. 2. Frequency distributions of twitch contraction times for motor units in the
cat gastrocnemius muscle (top) and the human tibialis anterior muscle (bottom).
The cat data (n¼ 117) were obtained by measuring the contraction time of the
twitch force elicited by applying single electric stimuli to the axons of single
motor units (Burke et al., 1973). The human data (n¼ 528) comprise spike-
triggered-average estimates of contraction times for motor units that were
activated during voluntary contractions in which the test motor unit discharged
action potentials at r 10 pps (Van Cutsem et al., 1997). The two histograms
indicate that the contraction times did not cluster into distinct groups of slow- and
fast-twitch motor units.
et al. (1973), the stimulus frequency was selected to minimize the
failure of muscle fiber activation and thereby to stress the
physiological processes distal to the muscle fiber action potential.
Two of the significant outcomes of this approach are the func-
tional relevance of the measure of fatigability and the interpreta-
tion that some motor units are fatigue resistant. One principle to
emerge in the literature on muscle fatigue is the concept that the
underlying mechanisms vary with the demands of the task being
performed (Enoka and Stuart, 1992). Although the fatigability
induced by some tasks can be attributed to mechanisms distal to
muscle fiber action potentials, many tasks are limited by impair-
ments in muscle activation (Enoka and Duchateau, 2008). Indeed,
it is likely that limitations in activities of daily living are more
related to activation issues rather than the capacity of muscle to
develop force or power.
The second issue involves the fatigability of fatigue-resistant
motor units. When human volunteers performed a series of 25
ramp-up and ramp-down contractions that were minimally
fatigable (9% decline in maximal voluntary contraction [MVC]
force), Farina et al. (2009) found that the adjustments were
greatest for the first recruited motor units. As the target force
for the ramp contractions was only 10% MVC force, the involved
429
motor units were presumably fatigue resistant based on the
motor unit-typing scheme. Nonetheless, the motor units that
discharged action potentials for a greater proportion of the
duration for the initial ramp contractions experienced greater
declines in conduction velocity of muscle fiber action potentials
and the proportion of the task during which action potentials
least active motor units declined over the course of the 25 ramp
contractions, which indicates that the earlier recruited motor
progressed. These results indicate that low-threshold motor units,
presumably fatigue resistant, are actually fatigable during volun-
Taken together, these findings indicate that motor units
cannot be distinguished on the basis of twitch contraction times
110 of imposed sequences of electrical stimulation does not provide a
functionally useful metric of motor unit properties (Bigland-
Ritchie et al., 1998). It is time that we abandoned the concept of
motor unit types.
3. Myoelectric manifestations of fatigue
Voluntary contractions arise from the activation of a motor
neuron pool by synaptic inputs that are provided by descending
pathways, spinal interneurons, and peripheral afferent feedback.
After these inputs have been integrated by the motor neurons, the
activation signal sent from the spinal cord to muscle comprises
trains of action potentials for the motor units that have been
recruited for the action. The net motor unit activity, therefore, is
100 the spinal cord and can be related to the force exerted by the
muscle. Consequently, selected properties of EMG signals have
been used as an index of the fatigue-associated adjustments in
motor unit activity during sustained and repetitive contractions
(Molinari et al., 2006; Rainoldi et al., 2008; Watanabe and Akima,
2010). Such an interpretation, however, assumes a stable associa-
tion between the EMG signal and muscle force. Unfortunately,
muscle force can vary due to changes in motor unit twitch force
and contraction velocity without changing EMG amplitude
(Carpentier et al., 2001; Fuglevand et al., 1999; Thomas et al.,
1991), and EMG amplitude can be modulated by changes in the
shapes and propagation velocity of the motor unit action poten-
tials (Dimitrova and Dimitrov, 2003; Keenan et al., 2005) without
a concurrent change in muscle force.
3.1. EMG–force relation
To estimate the extent to which the EMG–force relation can be
changed during fatiguing contractions, Dideriksen and colleagues
developed a computational model (Fig. 3) based on an earlier
model of motor unit recruitment and rate coding (Fuglevand et al.
1993a). The adjustments in motor unit activity during the
fatiguing contractions were implemented with a compartment-
model approach as functions of the metabolite concentration
within each muscle fiber and in the extracellular space
(Dideriksen et al., 2010a). The simulated concentrations were
related to the decrease in conduction velocity of muscle fiber
action potentials, increase in inhibitory afferent feedback, decline
in twitch-force amplitude, and the progressive inability of the
CNS to produce an output that matched the target force. The
model (Fig. 3) is able to reproduce a number of the adjustments
observed experimentally during fatiguing contractions: depres-
sion of motor unit action potentials, the relation between target
force and time to task failure, size-dependent reductions in motor
430 R.M. Enoka / Journal of Biomechanics 45 (2012) 427–433

Descending
drive EMG
model

n = 120

Motor neuron
Muscle
pool Neural drive

n = 40,500
Afferent Force
feedback model

Fig. 3. The fatigue-related changes in the association between surface EMG amplitude and muscle force was examined with a computational model of motor unit
recruitment and rate coding in which the time-varying accumulation of metabolites in a compartment model provided feedback that adjusted the conduction velocity of
muscle fiber action potentials, inhibitory afferent feedback to the motor neuron pool, twitch-force amplitude, and the ability of the CNS to produce an output that matched
the target force (Dideriksen et al., 2010a, b). The adjustments during the simulated contractions involved changes in the number of activated motor units and the rates at
which they discharged action potentials (neural drive) and the total number of muscle fiber action potentials (muscle activation).

unit tetanic force, recruitment of motor units and decreases in
discharge rate, and activity-dependent declines in recruitment
threshold (Dideriksen et al., 2010a).
The model was used to simulate changes in EMG amplitude
and muscle force for three representative fatigue protocols:
(1) repeated ramp contractions to 100% of the maximal voluntary
contraction (MVC) force; (2) contractions sustained at two target
forces (30 and 60% MVC force) for a similar force-time integral
and then followed by ramp contractions; and (3) contractions
sustained at five target forces (20–80% MVC force) for longer than
the time to task failure (Dideriksen et al., 2010b). The association
between EMG amplitude and muscle force varied across the
protocols being bounded by the relation in the absence of the
fatigue and the adjustments observed when the simulated con-
tractions were sustained longer than task failure (Fig. 4). Across
the three representative fatigue protocols, therefore, the same
muscle force was associated with EMG amplitudes that differed
by up to 25% of the MVC value. A similar fatigue-related
dissociation with muscle force presumably emerges for other
properties of the EMG signal.
3.2. Fatigue induced depression of EMG amplitude
Dideriksen and colleagues subsequently used the computa-
tional model to determine the adjustments that are responsible
for the depression of EMG amplitude when a low-force isometric
contraction is sustained for as long as possible (Fuglevand et al.,
1993b). The approach was to simulate the adjustments in motor
unit activity that were required to sustain isometric contractions
at target forces of 20, 40, and 60% of MVC force for as long as
possible (Dideriksen et al., 2011). The depression of EMG ampli-
tude at task failure of long-duration contractions was mainly
caused by a decrease in muscle activation (number of muscle
fiber action potentials) due to a decrease in net synaptic input to
motor neurons, with less of an influence due to changes in the
shapes of motor unit action potentials and no contribution by
amplitude cancellation (Dideriksen et al., 2011). Significantly,
EMG amplitude during the simulated fatiguing contractions was
related to the number of muscle fiber action potentials (muscle
activation), but not consistently to the number of motor unit
action potentials (neural drive to the muscle). However, the slope
of the relation between EMG amplitude and muscle activation
was inversely related to mean muscle fiber conduction velocity.
The simulations by Dideriksen and colleagues indicate that the
relation between surface EMG amplitude and muscle force is not
constant during fatiguing contractions and that neither the
amount of muscle activation (number of muscle fiber action
potentials) nor the level of neural drive (number of motor unit
action potentials) can be reliably estimated from EMG amplitude
during fatiguing contractions.
4. Fatigue and fatigability
The fatigability of muscle is classically quantified as the
decline in MVC force after performing some form of demanding
physical activity (Enoka and Duchateau, 2008; Gandevia, 2001).
At least since the observations of Mosso (1906), it has been
known that the mechanisms responsible for the decline in MVC
force depend on the characteristics of the task that induced the
fatigue (Enoka and Stuart, 1992). Although the evidence clearly
indicates that fatigue cannot be attributed to any single mechan-
ism, it has proven difficult to ascribe the fatigue experienced
under different conditions to specific sets of mechanisms.
Despite the emphasis in the fatigue literature on the impair-
ment of physiological processes distal to the muscle fiber action
potential (Ferreira and Reid, 2008; Fitts, 2008; Place et al., 2010),
even actions that involve only a modest demand are accompanied
by adjustments in motor unit activity (Farina et al., 2009). The
magnitude of these adjustments is underscored by the influence
of load compliance on the duration that a submaximal, isometric
contraction can be sustained. For example, the duration that an
individual can pull up against a rigid restraint with the elbow
flexor muscles to match a target of 20% MVC force (force control)
is approximately twice as long as when the same net muscle
torque is used to support an inertial load while keeping the arm in
the same position (position control) (Hunter et al., 2002). The
briefer time to failure during position control was associated with
greater adjustments in motor unit activity (mean discharge rate,
variability in discharge times, and recruitment), undoubtedly due
to differences in the synaptic input received by the motor neuron
pool (Klass et al., 2008; Mottram et al., 2005; Rudroff et al., 2010).
In addition to maintaining the requisite muscle force, however,
the fatigue-related adjustments in motor unit activation also
underlie the sensations that accompany fatiguing contractions
and there is typically a strong association between fatigability
 R.M. Enoka / Journal of Biomechanics 45 (2012) 427–433 431

100 1 protein, IL-6, TNF-alpha, thyroid-stimulating hormone, thyrox-

ine). However, those individuals classified as being fatigued (71
1 men and 172 women) reported worse sleep quality, lower self
2 80%
EMG amplitude (% MVC)

80 evaluations of health, and lower physical function scores, as

65% indicated by handgrip strength, the Short Physical Performance
Battery, average 400 m walking speed, the ability to walk 400 m,
4
and self-reported difficulties with activities of daily living and
60 50%
instrumented activities of daily living.
Measures of fatigability in older adults have produced equally
divergent observations, with some studies finding that old adults
8
40 35% are less fatigable than young adults and other studies reporting
the converse result. For example (Fig. 5), the time to task failure
20% for an isometric contraction (force control) with the elbow flexor
20 muscles sustained at 20% MVC for as long as possible was
0 20 40 60 80 22.677.4 min for old men (67–76 yrs; n ¼8) and 13.075.2 min
Force (% MVC) for strength-matched young men (18–31 yrs; n ¼8), even though
both groups experienced a similar decrease in MVC torque
100 (–31.4710.6%) at task failure (Hunter et al., 2005). In contrast,
Task failure the decline in the maximal torque exerted by the ankle dorsi-
flexor muscles during 5 sets of 30 maximal isokinetic (301/s)
No fatigue
80 contractions was greater for old adults (72–87 yrs; n¼16) com-
EMG amplitude (% MVC)

pared with young adults (22–47 yrs; n¼ 16) even though the
initial maximal torque (38.373.1 N m vs. 28.6 71.3 N m) was
60 greater for the young subjects (Baudry et al., 2007). The decrease

40

20

0
0 20 40 60 80 100
Force (% MVC)

Fig. 4. The relation between the amplitude of the surface EMG signal and muscle
force changes during fatiguing contractions. The changes in the associations
between EMG amplitude and muscle force were examined with the computational
model shown in Fig. 3 by simulating three representative fatigue protocols
(Dideriksen et al. 2010b). One protocol (top) involved simulating contractions at
five submaximal target forces (20, 35, 50, 65, and 85% of MVC force) beyond the
point in time when the muscle can achieve the target force and both EMG
amplitude and muscle force began to decline. The numbers on each trace (1, 2, 4,
or 8) indicate a time point (min) in each simulated contraction. The bottom panel
summarizes the simulated relations between EMG amplitude and muscle force
from the three fatigue protocols. The blue line indicates the relation in the absence 100
of fatigue, whereas the red line denotes the limit observed during the simulated
Young
protocol shown in the top panel. The other two protocols produced associations
Maximal torque (% initial)

that ranged between these two boundaries. The results indicate that EMG 90 Old
amplitude was not uniquely related to muscle force during the simulated fatiguing
contractions.
80

and rating of perceived exertion (Enoka and Duchateau, 2008; 70

Gandevia, 2001; Mosso, 1906).
In the clinical literature, however, perceptions of fatigue are
often unrelated to measures of fatigability (Zwarts et al., 2008).
Indeed, individuals can report being fatigued even in a rested
state. For example, 243 of the 812 participants (65–102 yrs; mean
75 yrs) enrolled in the InCHIANTI study reported being fatigued
in the absence of any physical activity (Vestergaard et al., 2009).
Fatigue was conceptualized as ‘‘the awareness of a decreased
capacity for physical and/or mental activity due to an imbalance
in the availability, utilization, and/or restoration of resources
needed to perform activity’’ and was assessed by two questions
from the Center for Epidemiologic Studies–Depression Scale.
There were no differences between the two groups of participants
(fatigued and not fatigued) in most clinical conditions and the
levels for a range of biomarkers (e.g., hemoglobin, C-reactive
60
50
0 30 60 90 120 150
Repetitions (5 x 30)
Fig. 5. The relative fatigability of young and old adults differs for isometric (top)
and anisometric (bottom) contractions. The results in the top graph comprise the
times to failure for strength-matched young and old men when they sustained an
isometric contraction with the elbow flexor muscles at 20% MVC torque for as long
as possible (Hunter et al., 2005). The time to failure of the old men (22.6 7 7.4 min)
was longer than that for the young men (13.0 7 5.2 min), indicating that they were
less fatigable with this task. In contrast, the rate of decline in maximal torque
during 5 sets of 30 repetitions with maximal lengthening contractions with the
ankle dorsiflexor muscles was greater for old adults (Baudry et al., 2007).
432 R.M. Enoka / Journal of Biomechanics 45 (2012) 427–433
in maximal torque was greater for old adults than for young
adults after both shortening (–50.2% vs. –40.9%, respectively) and
lengthening (–42.1% vs. –27.1%) contractions. Consistent with
these two examples, a subsequent meta-analysis concluded that
old adults are generally less fatigable than young adults when the
protocol involves isometric contractions, whereas the old adults
are more fatigable when the protocol involves anisometric con-
tractions (Christie et al., 2011).
Although the different observations on the relative fatigability
of old adults can presumably be explained by the task-dependent
demands on the underlying physiological processes, the critical
issue is which tasks provide insight on the mechanisms respon-
sible for the fatigue experienced by old adults. The lack of
progress on this question suggests that it cannot be resolved by
standard reductionist approaches, but that it requires a more
global perspective (Knicker et al., 2011; Noakes et al., 2005; Nybo,
2008; Tucker, 2009). Such an approach needs to include
both perceptions of fatigue and the mechanisms that define
fatigability, where the perceptions include both homeostatic
and psychological factors. The homeostatic factors involve those
physiological processes required to maintain the energetic and
physical integrity of the body and the psychological factors
include expectations, arousal, motivation, and mood. The
mechanisms that establish levels of fatigability correspond to
the physiological processes that are impaired during the perfor-
mance of various types of motor tasks. The fatigue experienced by
an individual, therefore, depends on both perceptions of fatigue
and the level of fatigability. With such a construct it is possible to
align the fatigue experienced by individuals in the absence of
exercise, such as older adults (Alexander et al., 2010; Eldadah,
2010; Vestergaard et al., 2009) and those afflicted with various
disorders and diseases (Ament and Verkerke, 2009; Féasson et al.,
2006; Zwarts et al., 2008), with that exercise-associated changes
in fatigue.
As an example of this approach, Steens et al. (in press)
compared the fatigue reported by 20 individuals (20–58 yrs; 13
women) with multiple sclerosis to that of 20 sex- and age-
matched control subjects. Although the most common complaints
of individuals with multiple sclerosis are fatigue and limitations
in walking, symptomatic fatigue is often not associated with
impairments in motor function (Hadjimichael et al., 2008;
Patrick et al., 2009; Putzki et al., 2008). The patients in the study
by Steens et al. (in press) were moderately impaired with
Extended Disability Severity Scores (EDSS) of r5.5 (can walk
without assistance) and a sense of fatigue (Fatigue Severity Scale
[FSS] questionnaire) that was greater (5.370.9) than that for the
control subjects (2.970.6). The purpose of the study was to
examine the associations between the sense of fatigue (FSS
scores) and the fatigability of a hand muscle. The patients were
weaker than the control subjects, but the decline in MVC force
during a sustained 2 min contraction did not differ between the
two groups of subjects. Nonetheless, the sense of fatigue reported
by the patients, but not the control subjects, was predicted
(r2 ¼ 0.45) by the decline in MVC force (r¼ 0.37), which was a
measure of fatigability, and the initial MVC force (r ¼0.35).
However, the sense of fatigue was more strongly predicted
(r2 ¼ 0.64) by the combination of depression score (Hospital
Anxiety and Depression Scale [HADS] questionnaire) and the
normalized MVC force. The sense of fatigue experienced by the
patients, therefore, was more strongly related to psychological
factors and the strength of a hand muscle than it was to any
measure of fatigability for the hand muscle.
Motivated by the lack of progress on translating observations
on the mechanisms underlying laboratory measures of fatigability
to the impairments experienced by individuals during activities of
daily living, it seems necessary to question our current
conceptualization of fatigue. A more global perspective seems
appropriate in which fatigue can be influenced by both percep-
tions of fatigue and the mechanisms the establish levels of
fatigability. Perhaps such a framework can provide a more
effective foundation from which to establish the functional
relevance of the diverse observations on fatigue.
Conflict of interest statement
None declared.
Acknowledgements
Among the many colleagues with whom I have been privileged
to work, I would like to acknowledge the significant contributions
of Professor Jacques Duchateau (Université Libre de Bruxelles)
and Professor Dario Farina (Georg-August University of
Göttingen) to the ideas and findings described in this paper.
References
Alexander, N.B., Taffet, G.E., Horne, F.M., Eldadah, B.A., Ferrucci, L., Nayfield, S.,
Studenski, S., 2010. Journal of the American Geriatric Society. Bedside-to-
bench conference: research agenda for idiopathic fatigue and aging 58,
96–975.
Ament, W., Verkerke, G.J., 2009. Exercise and fatigue. Sports Medicine 39,
389–422.
Baudry, S., Klass, M., Pasquet, B., Duchateau, J., 2007. Age-related fatigability of the
ankle dorsiflexor muscles during concentric and eccentric contractions.
European Journal of Applied Physiology 100, 515–525.
Bigland-Ritchie, B., Fuglevand, A.J., Thomas, C.K., 1998. Contractile properties of
human motor units: is man a cat? The Neuroscientist 4, 240–249.
Burke, R.E., Levine, D.N., Tsairis, P., Zajac III, F.E., 1973. Physiological types and
histochemical profiles in motor units of the cat gastrocnemius. Journal of
Physiology 234, 723–748.
Carp, J.S., Herchenroder, P.A., Chen, X.Y., Wolpaw, J.R., 1999. Sag during unfused
tetanic contractions in rat triceps surae motor units. Journal of Neurophysiol-
ogy 81, 2647–2661.
Carpentier, A., Duchateau, J., Hainaut, K., 2001. Motor unit behavior and contractile
changes during fatigue in human first dorsal interosseus. Journal of Physiology
534, 903–912.
Christie, A., Snook, E.M., Kent-Braun, J.A., 2011. Systematic review and meta-
analysis of skeletal muscle fatigue in old age. Medicine and Science in Sports
and Exercise 43, 568–577.
Delimata, J., 1996. Actes du Colloque Marey/Muybridge, pionniers du cinema-
recontre Beaune/Stanford: 19 mai 1995, Palais des congres, Beaune. Beaune,
Conseil regional de Bourgogne and Stanford, Université de Stanford.
Dideriksen, J.L., Enoka, R.M., Farina, D., 2011. Neuromuscular adjustments that
constrain submaximal EMG amplitude at task failure of sustained isometric
contractions. Journal of Applied Physiology 111, 485–494.
Dideriksen, J.L., Farina, D., Bækgaard, M., Enoka, R.M., 2010a. An integrative model
of motor unit activity during sustained submaximal contractions. Journal of
Applied Physiology 108, 1550–1562.
Dideriksen, J.L., Farina, D., Enoka, R.M., 2010b. Influence of fatigue on the
simulated relation between the amplitude of the surface electromyogram
and muscle force. Philosophical Transactions of the Royal Society A 368,
2765–2781.
Dimitrova, N.A., Dimitrov, G.V., 2003. Interpretation of EMG changes in fatigue:
facts, pitfalls, and fallacies. Journal of Electromyography and Kinesiology 13,
13–36.
Duchteau, J., Enoka, R.M., 2011. Human motor unit recordings: origins and insights
into the integrated motor system. Brain Research 1409, 42–61.
Eldadah, B.A., 2010. Fatigue and fatigability in older adults. Physical Medicine and
Rehabilitation 2, 406–413.
Enoka, R.M., Duchateau, J., 2008. What, why and how it influences muscle
function. Journal of Physiology 586, 11–23.
Enoka, R.M., Stuart, D.G., 1992. Neurobiology of muscle fatigue. Journal of Applied
Physiology 72, 1631–1648.
Farina, D., Holobar, A., Gazzoni, M., Zazula, D., Merletti, R., Enoka, R.M., 2009.
Adjustments differ among low-threshold motor units during intermittent,
isometric contractions. Journal of Neurophysiology 101, 350–359.
Féasson, L., Camdessanché, J.P., El Mandh, L., Calmeis, P., Millet, G.Y., 2006. Fatigue
and neuromuscular diseases. Annales of Réadaptation et de Médecine Physi-
que 49, 375–384.
Ferreira, L.F., Reid, M.B., 2008. Muscle-derived ROS and thiol regulation in muscle
fatigue. Journal of Applied Physiology 104, 853–860.
 R.M. Enoka / Journal of Biomechanics 45 (2012) 427–433
Fitts, R.H., 2008. The cross-bridge cycle and skeletal muscle fatigue. Journal of
Applied Physiology 104, 551–558.
Fuglevand, A.J., Macefield, V.G., Bigland-Ritchie, B., 1999. Force-frequency and
fatigue properties of motor units in muscles that control digits of the human
hand. Journal of Neurophysiology 81, 1718–1729.
Fuglevand, A.J., Winter, D.A., Patla, A.E., 1993a. Models of recruitment and rate
coding organization in motor-unit pools. Journal of Neurophysiology 70,
2470–2488.
Fuglevand, A.J., Zackowski, K.M., Huey, K.A., Enoka, R.M., 1993b. Impairment of
neuromuscular propagation during human fatiguing contractions at submax-
imal forces. Journal of Physiology 460, 549–572.
Gandevia, S.C., 2001. Spinal and supraspinal factors in human muscle fatigue.
Physiological Reviews 81, 1725–1789.
Hadjimichael, O., Vollmer, T., Oleen-Burkey, M., North American Research Com-
mittee on Multiple Sclerosis, 2008. Fatigue characteristics in multiple sclero-
sis: the North American Research Committee on Multiple Sclerosis
(NARCOMS) Survey. Health Quality of Life Outcomes 6, 100.
Hunter, S.K., Critchlow, A., Enoka, R.M., 2005. Muscle endurance is greater for old
men compared with strength-matched young men. Journal of Applied Phy-
siology 99, 890–897.
Hunter, S.K., Ryan, D.L., Ortega, J.D., Enoka, R.M., 2002. Task differences with the
same load torque alter the endurance time of submaximal fatiguing contrac-
tions in humans. Journal of Neurophysiology 88, 3087–3096.
Keenan, K.G., Farina, D., Maluf, K.S., Merletti, R., Enoka, R.M., 2005. Influence of
amplitude cancellation on the simulated surface electromyogram. Journal of
Applied Physiology 98, 120–131.
Klass, M., Lévénez, M., Enoka, R.M., Duchateau, J., 2008. Spinal mechanisms
contribute to differences in the time to failure of submaximal fatiguing
contractions performed with different loads. Journal of Neurophysiology 99,
1096–1104.
Knicker, A.J., Renshaw, I., Oldham, A.R.H., Cairns, S.P., 2011. Interactive processes
link the multiple causes of fatigue in sport competition. Sports Medicine 41,
307–328.
Molinari, F., Knaflitz, M., Bonato, P., Actis, M.V., 2006. Electrical manifestations of
muscle fatigue during concentric and eccentric isokinetic knee flexion-exten-
sion movements. IEEE Transactions on Biomedical Engineering 53, 1309–1316.
Mosso, A., 1906. Fatigue. (Translated by Drummond M, Drummond WB). Swan
Sonnenschein & Co. Ltd, London.
Mottram, C.J., Jakobi, J.M., Semmler, J.G., Enoka, R.M., 2005. Motor-unit activity
differs with load type during a fatiguing contraction. Journal of Neurophysiol-
ogy 93, 1381–1392.
Noakes St., T.D., Clair Gibson, A., Lambert, E.V., 2005. From catastrophe to
complexity: a novel model of integrative central neural regulation of effort
433
and fatigue during exercise in humans: summary and conclusions. British
Journal of Sports Medicine 39, 120–124.
Nybo, L., 2008. Hyperthermia and fatigue. Journal of Applied Physiology 104,
871–878.
Patrick, E., Christodoulou, C., Krupp, L.B., New York State MS Consortium, 2009.
Longitudinal correlates of fatigue in multiple sclerosis. Multiple Sclerosis 15,
258–261.
Place, N., Yamada, T., Bruton, J.D., Westerblad, H., 2010. Muscle fatigue: from
observations in humans to underlying mechanisms studied in intact single
muscle fibres. European Journal of Applied Physiology 110, 1–15.
Putzki, N., Katsarava, Z., Vago, S., Diener, H.C., Limmroth, V., 2008. Prevalence and
severity of multiple-sclerosis-associated fatigue in treated and untreated
patients. European Neurology 59, 136–142.
Rainoldi, A., Falla, D., Mellor, R., Bennell, K., Hodges, P., 2008. Myoelectric
manifestations of fatigue in vastus lateralis, medialis obliquus and medialis
longus muscles. Journal of Electromyogrophy and Kinesiology 18, 1032–1037.
Rudroff, T., Jordan, K., Enoka, J.A., Matthews, S.D., Baudry, S., Enoka, R.M., 2010.
Discharge of biceps brachii motor units is modulated by load compliance and
forearm posture. Experimental Brain Research 202, 111–120.
Sherrington, C.S., 1925. Remarks on some aspects of reflex inhibition. Proceedings
of the Royal Society of London B97, 19–45.
Steens, A., de Vries, A., Hemmen, J., Heersema, T., Heerings, M., Mauritis, N.,
Zijdewind, I. Fatigue perceived by multiple sclerosis patients is associated with
muscle fatigue. Neurorehabilitation and Neural Repair. In press.
Thomas, C.K., Johansson, R.S., Bigland-Ritchie, B., 1991. Attempts to physiologically
classify human thenar motor units. Journal of Neurophysiology 65,
1501–1508.
Tucker, R., 2009. The anticipatory regulation of performance: the physiological
basis for pacing strategies and the development of a perception-based model
for exercise performance. British Journal of Sports Medicine 43, 392–400.
Van Cutsem, M., Feiereisen, P., Duchateau, J., Hainaut, K., 1997. Mechanical
properties and behavior of motor units in the tibialis anterior during voluntary
contractions. Canadian Journal of Applied Physiology 22, 585–597.
Vestergaard, S., Nayfield, S.G., Patel, K.V., Eldadah, B., Cesari, M., Ferrucci, L.,
Ceresini, G., Guralnik, J.M., 2009. Fatigue in a representative population of
older persons and its association with functional impairment, functional
limitation, and disability. Journal of Gerontology A Biological Sciences Medical
Sciences 64A, 76–82.
Watanabe, K., Akima, H., 2010. Neuromuscular activation of vastus intermedius
muscle during fatiguing exercise. Journal of Electromyography and Kinesiol-
ogy 20, 661–666.
Zwarts, M.J., Bleijenberg, G., van Engelen, B.G., 2008. Clinical neurophysiology of
fatigue. Clinical Neurophysiology 119, 2–10.